DESCRIPTION: (Applicant's Abstract) This project will provide data to construct an initial theoretical model and make predictions about the long-term influences of siblings on substance use, with direct implications for future development of intervention and prevention programs for at-risk youth. Developmental models which have previously included child behavior, parents, and peers as predictors of later substance use will be extended to account for the effects of siblings on later behavior. Both an ecological model and a social learning perspective will be integrated to account for the relationship between qualitative aspects of the sibling relationship and later substance use. Youth at risk for substance abuse typically have a number of associated problems, including a history of antisocial behavior, peer relationship difficulties, and parents with limited family management skills. Previous research has highlighted the influence of both peers and parents in later substance use. In addition, several recent studies have found siblings to be major contributors to later substance use; however, the mechanisms by which siblings may influence later behavior is still unknown. The proposed project would involve adding several data collection procedures to a currently funded intervention trial which serves multi-ethnic children and families at risk for later substance use (called Project Alliance, funded by NIDA). The primary assessment would be direct observation of a sibling interaction task, which would be coded for conflict, positive support, deviant behavior, and substance use in the relationship. Self-reports of the sibling relationship would also be obtained from target youth, their siblings, and parents. Data would be collected over a 2-year period which will allow for the development of a short-term longitudinal model. Group comparisons by race and gender composition of the sibling dyad will also be conducted.